Grafton Sciences is building physical general intelligence, and this role is to build high-fidelity digital twins of robotic, electromechanical, and experimental systems to enable accurate predictive simulation and closed-loop interaction with RL, planning, and control stacks.
Requirements
- Strong experience building or calibrating digital twins, dynamic models, or data-driven physics models.
- Familiarity with system identification, time-series modeling, physical parameter estimation, and stability/fidelity considerations.
- Ability to blend physics, machine learning, and experimental data into robust predictive models.
- Comfortable working across ML, simulation tools, and physical hardware interfaces in a fast-moving research and engineering environment.
Responsibilities
- Develop model-identification pipelines, parameter fitting routines, and adaptive calibration systems for digital twins.
- Build ML-based dynamic models, multi-scale physics approximators, and hybrid simulation frameworks.
- Ensure twin fidelity, stability, and cross-version consistency as real systems change or new data arrives.
- Collaborate with simulation, RL, controls, and agent systems teams to integrate digital twins into learning and decision-making workflows.
- Design model-identification pipelines, calibration routines, dynamic-model learning systems, and multi-scale representations that enable accurate predictive simulation and closed-loop interaction with RL, planning, and control stacks.
Other
- Candidates who can demonstrate world-class excellence.
